By and large, U.S. media have spun the July employment report as more positive than negative. The 117,000 new payroll jobs created last month and the upward revisions for May and June were a relief after two months of very bad data. The downtick in the unemployment rate, although slight, was also welcome. One indicator that rarely makes the headlines told a different story, however. The employment-population ratio fell to a new low of 58.1 percent. What does it mean? Why should we care?
We should care, because the sinking employment-population ratio has big implications for the budget debate. The cuts-only faction of budget balancers are demanding a cap on federal government spending at 18% of GDP. They tout that as a level we lived with happily in the past, and should therefore be happy to live with in the future. The trouble is, the future isn’t going to be like the past. The smaller the fraction of the population that is working, the harder it becomes to put the country’s fiscal affairs in order. That becomes even clearer if we take a closer look at the reasons the ratio is falling.
The employment-population ratio depends, first, on the percentage of the population that is in the labor force (that is, either working or actively looking for work), and second, on the percentage of the labor force that is employed (that is, one minus the unemployment rate). Both of these numbers have a cyclical component. The unemployment rate rises during a downturn, and labor force participation falls when some people become discouraged by the lack of jobs and stop actively looking for work.
The cyclical nature of the employment-population ratio can be seen in the following chart, where recessions are shown by shaded vertical bars. The five recessions through 1991 have a consistent pattern. The employment ratio falls during the recession, and then, with a fairly short lag, recovers again. The pattern in the two recessions of the 2000s is strikingly different. In the more recent cases, the employment ratio falls during the recession, but at a rate close to a trend that is now moving downward. After the recession ends, the employment rate continues to fall with little or no visible recovery.
The behavior of the employment-population ratio during the last two recessions in part reflects the much slower recovery of the unemployment rate, familiar from a widely reproduced graphic supplied by Calculated Risk. That is only part of the story, though. There are also strong structural influences at work, which we can see by breaking the employment ratio down by population segment, as in the next chart.
In this chart, the ratio is shown separately for adult men, adult women, and teenagers. Each shows a different pattern. The percentage of adult men who are employed decreased fairy steadily from the 1950s through the 1970s, flattened out during the 1980s and 1990s, and then resumed its downward trend after 2000.
For most of the second half of the 20th century, however, the downward trend of men’s employment was more than offset by a long, steady increase in the share of employed women. The women’s employment ratio peaked in the late 1990s, flattened out, and has now begun to fall again.
Meanwhile, teenage employment, although more sensitive to cyclical influences than that of adults, showed little upward or downward trend from the 1950s through the 1990s. Since 2000, it has fallen precipitously, to the point that teenage workers have become almost a rarity. The July, 2011 data show labor force participation for teenagers of just 34 percent, and among them, an unemployment rate of 25 percent.
Several social and economic factors lie behind these trends, including changing men’s and women’s attitudes toward work and a changing teenage job market. In coming years, however, demographic factors will increasingly be the dominant influence, as shown by the next chart, which gives data on changing dependency ratios (percentages of the population not of working age).
Two features stand out in the chart:
First, we see that dependency ratios have not yet played a role in the decline of the employment-population ratio. After reaching an all-time low in the mid-1980s, the total dependency ratio changed little through 2010. It is only now starting to rise with the retirement of the first baby-boomers.
Second, when we look at the breakdown of the total ratio, we see that its composition in the future will be nothing like that of the equally high total ratios of the late 1950s and 1960s. At that time, the majority of dependents were children. They were cared for, in large part, by full-time moms, and were fed, clothed, and housed by dads with strong attachments to the labor force. In contrast, the dependents of the future will be retirees.
To some extent they will be cared for by their working-age children, who will mow their lawns and drive them to the mall. However, those children, sons and daughters alike, will have to stay tightly attached to the labor force in order to pay the taxes that will be needed to fund social security and Medicare benefits for their retired parents and grandparents.
What do these trends mean for the budget debate? They tell us that a cuts-only fiscal consolidation strategy with a cap of 18 percent on federal expenditures will not mean a return to the good old days of the 1950s and 1960s. Instead, holding total spending constant would mean very large cuts in social security and Medicare benefits per retiree, or very severe cuts to defense and non-defense discretionary spending, or both. Some cuts-only enthusiasts would, no doubt, welcome a return to Victorian levels of social services and an isolationist defense posture.
However, it is doubtful that cutting social security and Medicare benefits far below past levels will be acceptable to voters aged 65 and older, who, as the chart indicates, will soon (if they do not already) constitute an insurmountably strong voting bloc.
An alternative to a cuts-only budget strategy will have to include three elements.
First, it will of course have to include spending cuts. Subsidies to farmers, racetrack owners, and others will have to be re-examined. Foreign military commitments will have to be reconsidered. The retirement age may have to be indexed to longevity and Medicare may have to be means tested. Above all, there will have to be a serious discussion of how to reform the U.S. structure of health care costs so that they look more like those of other advanced countries that provide better care with lower outlays.
Second, there will have to be increases in revenues, if only because sufficient spending cuts will not, in the long run, be acceptable to voters.
Third, the revenue increases will have to be achieved through comprehensive tax reform, not just through increases in marginal tax rates. Reforms will have to pay special attention to features of the tax system that discourage job creation and labor-market participation. Marginal rates for payroll taxes, income taxes, and corporate profit taxes should be lowered, not raised.
However, there should be no illusion that doing so will generate needed revenues through some kind of Laffer-curve magic. Loopholes will have to be eliminated, including the really big ones, like the mortgage interest deduction, employer-paid health benefits, and charitable contributions. There should also be a shift from taxes that disincentivize work toward taxes on consumption, like a higher gasoline tax or a value-added tax.
Of course, everyone knows all this. Maybe policy makers are thinking that if they wait for things to get a little worse, it will be easier to act. Well, if we take the employment-population ratio as an indicator, things just did get a little worse. Maybe the time to act is now.